Multi-species richness of boreal agricultural landscapes: effects of climate, biotope, soil and geographical location

Aim To assess the relative importance of climate, biotope and soil variables as well as geographical location for the species richness of plants, butterflies, day‐active macromoths and wild bees in boreal agricultural landscapes. Location A total of 68 agricultural landscapes located in southern Finland. Methods Generalized linear mixed models were used to analyse the effects of environmental (climate, biotope and soil) and spatial (latitude and longitude) variables on species richness of four taxa in 136 study squares of 0.25 km². Using partial regression, the variation in species richness was decomposed into the purely environmental fraction; the spatially structured environmental fraction; and the purely spatial fraction, including variables retained in cubic trend surface regression. Results Species richness of all taxa was positively correlated with temperature. Species richness of plants and butterflies was also positively correlated with the heterogeneity of landscape. The extent of low‐intensity agricultural land and forest had a positive effect, and the extent of cultivated field a negative effect on the species richness of these taxa. The effect of soil characteristics on the number of observed species was negligible for all taxa. The greatest part of the explained variation for all taxa was accounted for by the pure effect of geographical location. To a somewhat lesser extent, the species richness of plants, butterflies and bees was also related to the effects of spatially structured environmental variables, and the species richness of macromoths to the effects of environmental variables. Main conclusions Multi‐species richness of boreal agricultural landscapes at the scale of 0.25 km² was associated with the heterogeneity of the local landscape. However, large‐scale geographical variation in species richness was also observed, which indicates the importance of climate and geographical location for the taxa studied. These results highlight the fact that, even on a landscape scale, geographical factors should be taken into account in biodiversity studies. Heterogeneous agricultural landscapes, including forest and non‐crop biotopes, should be preserved or restored to maintain biodiversity.     

Plant functional group composition and large‐scale species richness in European agricultural landscapes

Question: Which are the plant functional groups responding most clearly to agricultural disturbances? Which are the relative roles of habitat availability, landscape configuration and agricultural land use intensity in affecting the functional composition and diversity of vascular plants in agricultural landscapes? Location: 25 agricultural landscape areas in seven European countries. Methods: We examined the plant species richness and abundance in 4 km × 4 km landscape study sites. The plant functional group classification was derived from the BIOLFLOR database. Factorial decomposition of functional groups was applied. Results: Natural habitat availability and low land use intensity supported the abundance and richness of perennials, sedges, pteridophytes and high nature quality indicator species. The abundance of clonal species, C and S strategists was also correlated with habitat area. An increasing density of field edges explained a decrease in richness of high nature quality species and an increase in richness of annual graminoids. Intensive agriculture enhanced the richness of annuals and low nature quality species. Conclusions: Habitat patch availability and habitat quality are the main drivers of functional group composition and plant species richness in European agricultural landscapes. Linear elements do not compensate for the loss of habitats, as they mostly support disturbance tolerant generalist species. In order to conserve vascular plant species diversity in agricultural landscapes, the protection and enlargement of existing patches of (semi‐) natural habitats appears to be more effective than relying on the rescue effect of linear elements. This should be done in combination with appropriate agricultural management techniques to limit the effect of agrochemicals to the fields.     

Threat spots and environmental determinants of red-listed plant,butterfly and bird species in boreal agricultural environments

The aims of this study were (1) to examine the geographic distribution of red-listed species of agricultural environments and identify their national threat spots (areas with high diversity of threatened species) in Finland and (2) to determine the main environmental variables related to the richness and occurrence patterns of red-listed species. Atlas data of 21 plant, 17 butterfly and 11 bird species recorded using 10 km grid squares were employed in the study. Generalized additive models (GAMs) were constructed separately for species richness and occurrence of individual species of the three species groups using climate and land cover predictor variables. The predictive accuracy of models, as measured using correlation between the observed and predicted values and AUC statistics, was generally good. Temperature-related variables were the most important determinants of species richness and occurrence of all three taxa. In addition, land cover variables had a strong effect on the distribution of species. Plants and butterflies were positively related to the cover of grasslands and birds to small-scale agricultural mosaic as well as to arable land. Spatial coincidence of threat spots of plants, butterflies and birds was limited, which emphasizes the importance of considering the potentially contrasting environmental requirements of different taxa in conservation planning. Further, it is obvious that the maintenance of various non-crop habitats and heterogeneous agricultural landscapes has an essential role in the preservation of red-listed species of boreal rural environments.     

After the hotspots are gone: Land use history and grassland plant species diversity in a strongly transformed agricultural landscape

Question: We asked how landscape configuration and present management influence plant species richness and abundance of habitat specialists in grasslands in a ‘modern’(much exploited and transformed) agricultural Swedish landscape. Location: Selaön, south‐eastern Sweden (59°24’ N, 17°10’ E). Methods: Present and past (150 and 50 years ago) landscape pattern was analysed in a 25 km² area. Species richness was investigated in 63 different grassland patches; grazed and abandoned semi‐natural grasslands, and grazed ex‐arable fields. Influence of landscape variables; area, past and present grassland connectivity, present management on total species richness, density and abundance of 25 grassland specialists was analysed. Results: Semi‐natural grasslands (permanent unfertilised pastures or meadows formed by traditional agricultural methods) had declined from 60% 150 years ago to 5% today. There was a significant decline in species richness and density in abandoned semi‐natural grasslands. Total species richness was influenced by present management, size and connectivity to present and past grassland pattern. Landscape variables did not influence species density in grazed semi‐natural grassland suggesting that maintained grazing management makes grassland patches independent of landscape context. The abundance of 16 grassland specialists was mainly influenced by management and to some extent also by landscape variables. Conclusion: Although species richness pattern reflect management and to some extent landscape variables, the response of individual species may be idiosyncratic. The historical signal from past landscapes is weak on present‐day species richness in highly transformed, agricultural landscapes. Generalizations of historical legacies on species diversity in grasslands should consider also highly transformed landscapes and not only landscapes with a high amount of diversity hotspots left.     

Species richness of vascular plantsin the subarctic landscape of northern Finland:modelling relationships to the environment

This paper presents models based on empirical data which can be used to predict the patterns of species richness of vascular plants at the poorly explored mesoscale. Using generalized linear modelling, multiple regression models of species richness in the Kevo Nature Reserve, North Finland, are built with a training set of 257 grid squares and 33 environmental variables. We validated the accuracy of the derived models with an independent test set of 100 grid squares. Two different modelling approaches are used: one where species richness is treated straightforwardly as the response variable, and another where it is tentatively stratified into two groups according to taxon types, i.e. alpine taxa versus wide-spread and silvine (forest) taxa. However, the latter approach only marginally improved the accuracy of the predictions of total number of species. Linear altitudinal variables were among the best predictors of vascular plant richness at the mesoscale. As variables involving altitude are crude surrogates for energy-related factors, the results support the available energy hypothesis and advocate its significance in richness-environment relationships. Other important predictors of species richness included length of rivers and brooks, abundance of cliff walls, occurrences of steep-sided gorges and valleys, and relative abundance of gabbro in bedrock. However, the accuracy of the predictions in the derived models is relatively modest. This points towards the necessity of field work as a final guarantee to identify local hotspots of vascular plant species in a subarctic landscape. This revised version was published online in November 2006 with corrections to the Cover Date.     

Local diversity of arable weeds increases with landscape complexity

Patterns of plant diversity are often related to local site conditions and to competitive interactions, but landscape context may also be important for local plant species richness. This is shown here by analysing the relationship between landscape complexity and local species richness of arable weeds in wheat fields. The fields were located in 18 landscapes characterised by a gradient in landscape complexity from structurally complex to structurally simple (39–94% arable land). We quantified local site conditions, field management intensity and landscape characteristics, and used principle component analyses to ordinate the environmental variables. The percentage of arable land was negatively correlated with perimeter–area ratio, habitat-type diversity and topographical heterogeneity, but landscape characteristics did not correlate with local site conditions and field management intensity. The number of plant species was mainly related to landscape characteristics and to a lesser extent to field management intensity (nitrogen fertilisation), whereas local soil characteristics did not contribute to the explanation of arable weed richness. In a geographic scale analysis using circular landscape sectors ranging from 1 km up to 5 km diameter, the predictive power of landscape complexity for local plant species richness was strongest at 2 km indicating a scale-dependent relationship between landscape context and plant species richness. Our results support the hypothesis that local plant species richness in arable fields is greatly influenced by processes operating at the landscape scale. Seed rain from ruderal source habitats and disturbed edges may be the most important underlying process.     

Habitat quality and surrounding landscape structures influence wild bee occurrence in perennial wildflower strips

Perennial wildflower strips (WFS) are known to have positive effects on wild bees in intensively used agricultural landscapes. Little knowledge exists, however, about the drivers of wild bee occurrence and if Red List species also profit from this agri-environmental scheme (AES). Therefore, we studied wild bees on transects along 20 four- to five-year-old WFS and in 10 cereal fields without AES (CONTROL sites) in differently structured landscapes across Saxony-Anhalt (Germany). In addition to local site parameters, we measured parameters of landscape structure in a 1 km radius of the WFS and CONTROL sites. The overall species richness of wild bees (125 species in total, 23 on average), including numerous specialist and Red List species, indicates a high attractiveness of perennial WFS sown with 30 native forbs. In CONTROL fields, 11 bee species (on average only one) were found.The species richness and abundance of wild bees were positively affected by local site conditions of the WFS and CONTROL sites, such as the overall number of sown and spontaneous forbs, the amount of flower rewards of sown forbs available to pollinators (Pollinator Feeding Index), and negatively by the cover of grasses. Therefore, seed mixtures of future AES should comprise a high diversity of wildflower species relevant as pollen sources for wild bees. The share of Red List wild bee species was strongly influenced by the landscape context and increased e.g. with Shannon landscape diversity and the availability of non-forest woody habitats and water bodies in the 1 km surroundings. These results suggest that besides the establishment of high-diversity WFS, semi-natural habitat structures have to be promoted to preserve rare wild bees especially in structurally simple agricultural landscapes.     

Multiscale landscape genomic models to detect signatures of selection in the alpine plant Biscutella laevigata

Plant species are known to adapt locally to their environment, particularly in mountainous areas where conditions can vary drastically over short distances. The climate of such landscapes being largely influenced by topography, using fine‐scale models to evaluate environmental heterogeneity may help detecting adaptation to micro‐habitats. Here, we applied a multiscale landscape genomic approach to detect evidence of local adaptation in the alpine plant Biscutella laevigata. The two gene pools identified, experiencing limited gene flow along a 1‐km ridge, were different in regard to several habitat features derived from a very high resolution (VHR) digital elevation model (DEM). A correlative approach detected signatures of selection along environmental gradients such as altitude, wind exposure, and solar radiation, indicating adaptive pressures likely driven by fine‐scale topography. Using a large panel of DEM‐derived variables as ecologically relevant proxies, our results highlighted the critical role of spatial resolution. These high‐resolution multiscale variables indeed indicate that the robustness of associations between genetic loci and environmental features depends on spatial parameters that are poorly documented. We argue that the scale issue is critical in landscape genomics and that multiscale ecological variables are key to improve our understanding of local adaptation in highly heterogeneous landscapes.     

Do biogeographic parameters matter? Plant species richness and distribution of macrophytes in relation to area and isolation of ponds in NW Polish agricultural landscape

Effects of pond size and isolation on total vascular plant species richness and number of obligate wetland species were compared. Subsequently, the potential for the presence of spatial patterns in wetland species distribution among ponds in an agricultural landscape was explored. Relationships between species richness and two main biogeographic parameters were analysed using simple and multiple linearised regression models. Spatial patterns were looked for by means of analyses carried out with the R CRAN software (join-count statistics). Simple regression analyses performed on the regional scale (n = 50) revealed the significance of the effect of pond size only (r = 0.46 for total plant species richness and r = 0.28 for wetland species richness vs. pond area). Further analyses conducted on the local scale identified the best multiple regression models in the largest pond cluster (n = 20); the models showed statistical significance of relationships between the species richness and both independent variables (r = 0.80 for total plant species richness and r = 0.70 for wetland species richness vs. pond area and isolation, including mean distance to the nearest ten ponds). Spatial analyses were performed for 26 obligate wetland species selected from 149 species recorded in all the 50 ponds. Exploratory spatial data analysis revealed the presence of significant positive spatial autocorrelation in the distribution of 8 species. In such cases, it is possible to reject the random distribution hypothesis, which justifies exploration of spatial regimes. In practice, correct spatial model specifications may have implications for predicting species occurrences under changing environmental conditions, e.g. changes in the number of ponds.     

Short Rotation Coppice (SRC) Plantations Provide Additional Habitats for Vascular Plant Species in Agricultural Mosaic Landscapes

Increasing loss of biodiversity in agricultural landscapes is often debated in the bioenergy context, especially with respect to non-traditional crops that can be grown for energy production in the future. As promising renewable energy source and additional landscape element, the potential role of short rotation coppice (SRC) plantations to biodiversity is of great interest. We studied plant species richness in eight landscapes (225 km²) containing willow and poplar SRC plantations (1,600 m²) in Sweden and Germany, and the related SRC α-diversity to species richness in the landscapes (γ-diversity). Using matrix variables, spatial analyses of SRC plantations and landscapes were performed to explain the contribution of SRC α-diversity to γ-diversity. In accordance with the mosaic concept, multiple regression analyses revealed number of habitat types as a significant predictor for species richness: the higher the habitat type number, the higher the γ-diversity and the lower the proportion of SRC plantation α-diversity to γ-diversity. SRC plantation α-diversity was 6.9 % (±1.7 % SD) of species richness on the landscape scale. The contribution of SRC plantations increased with decreasing γ-diversity. SRC plantations were dominated more by species adapted to frequent disturbances and anthropo-zoogenic impacts than surrounding landscapes. We conclude that by providing habitats for plants with different requirements, SRC α-diversity has a significant share on γ-diversity in rural areas and can promote diversity in landscapes with low habitat heterogeneity and low species pools. However, plant diversity enrichment is mainly due to additional species typically present in disturbed and anthropogenic environments.     

BIODIVERSITY RESEARCH: Diversity and distribution of macroinvertebrates in lentic habitats in massively altered landscapes in south‐eastern Australia

Aim We investigated whether faunas of lentic macroinvertebrates differed among two landscape types: (1) those that are largely covered in forests (presumed to be in a more pre‐human‐impact condition) and (2) those that are completely cleared for agricultural exploitation (massively altered). Location Five pairs of landscapes (each pair referred to as a region) – one of each landscape type – across a 30,000 km² region of north‐central Victoria, Australia. Methods Each individual waterbody was surveyed three times (austral spring 2006, autumn 2007, and spring 2007) for invertebrates. Waterbodies were characterized by measurements of static (e.g. abutting vegetation cover) and labile (e.g. pH) variables. Data were analysed using hierarchical Bayesian models of species richness, α‐ and β‐diversities and functional feeding groups. Assemblage composition was related to landscape and in‐waterbody characteristics. Results Neither measured, nor asymptotic estimates of, species richness differed among landscape types, notwithstanding consistent differences in in‐waterbody habitat characteristics among waterbodies in the two landscape types. There were no discernible differences in patterns of α‐ and β‐diversities at landscape scales relating to landscape type. Habitat diversity of waterbodies at the landscape scale did not affect β‐diversity, although distinct waterbodies within landscapes tended to have more distinct faunas. Main conclusions The lentic macroinvertebrate faunas are relatively homogeneous over the entire region, with little differentiation between wooded and cleared landscapes. The regional fauna may be a homogenized subset of native species, possibly arising from the huge numerical predominance of lentic habitats in agricultural landscapes producing ‘spill‐over’ effects into forested landscapes. Of taxa more frequently found in one or other landscape type, trophic group diversity was greater in forested landscapes.     

The Role of Landscape Patterns of Habitat Types on Plant Species Diversity of a Tropical Forest in Mexico

The relationships among landscape characteristics and plant diversity in tropical forests may be used to predict biodiversity. To identify and characterize them, the number of species, as well as Shannon and Simpson diversity indices were calculated from 157 sampling quadrats (17,941 individuals sampled) while the vegetation classes were obtained from multi-spectral satellite image classification in four landscapes located in the southeast of Quintana Roo, Mexico. The mean number of species of trees, shrubs and vines as well as the mean value of the total number of species and the other two diversity indices were calculated for four vegetation classes in every one of the four landscapes. In addition, the relationships between landscape patterns metrics of patch types and diversity indices were explored. The multiple statistical analyses revealed significant predictor variables for the three diversity indices. Moreover, the shape, similarity and edge contrast metrics of patch types might serve as useful indicators for the number of species and the other two diversity variables at the landscape scale. Although the association between the three diversity indices and patch types metrics showed similar behavior, some differences were appreciated. The Shannon diversity index, with its greater sensitivity to rare species, should be considered as having a greater importance in interpretation analysis than Simpson index.     

Landscape genetics of a tropical rescue pollinator

Pollination services are increasingly threatened by the loss and modification of natural habitats, posing a risk to the maintenance of both native plant biodiversity and agricultural production. In order to safeguard pollination services, it is essential to examine the impacts of habitat degradation on the population dynamics of key pollinators and identify potential “rescue pollinators” capable of persisting in these human-altered landscapes. Using a landscape genetic approach, we assessed the impact of landscape structure on genetic differentiation in the widely-distributed tropical stingless bee Trigona spinipes (Apidae: Meliponini) across agricultural landscape mosaics composed of coffee plantations and Atlantic forest fragments in southeastern Brazil. We genotyped 115 bees at 16 specific and highly polymorphic microsatellite loci, developed using next-generation sequencing. Our results reveal that T. spinipes is capable of dispersing across remarkably long distances, as we did not find genetic differentiation across a 200 km range, nor fine-scale spatial genetic structure. Furthermore, gene flow was not affected by forest cover, land cover, or elevation, indicating that reproductive individuals are able to disperse well through agricultural landscapes and across altitudinal gradients. We also found evidence of a recent population expansion, suggesting that this opportunistic stingless bee is capable of colonizing degraded habitats. Our results thus suggest that T. spinipes can persist in heavily-altered landscapes and can be regarded as a rescue pollinator, potentially compensating for the decline of other native pollinators in degraded tropical landscapes.     

Coccinellid response to landscape composition and configuration

相似文献   

Noncrop features and heterogeneity mediate overwintering bird diversity in agricultural landscapes of southwest China

Farmland birds are of conservation concerns around the world. In China, conservation management has focused primarily on natural habitats, whereas little attention has been given to agricultural landscapes. Although agricultural land use is intensive in China, environmental heterogeneity can be highly variable in some regions due to variations in crop and noncrop elements within a landscape. We examined how noncrop heterogeneity, crop heterogeneity, and noncrop features (noncrop vegetation and water body such as open water) influenced species richness and abundance of all birds as well as three functional groups (woodland species, agricultural land species, and agricultural wetland species) in the paddy‐dominated landscapes of Erhai water basin situated in northwest Yunnan, China. Birds, crop, and noncrop vegetation surveys in twenty 1 km × 1 km landscape plots were conducted during the winter season (from 2014 to 2015). The results revealed that bird community compositions were best explained by amounts of noncrop vegetation and compositional heterogeneity of noncrop habitat (Shannon–Wiener index). Both variables also had a positive effect on richness and abundance of woodland species. Richness of agricultural wetland species increased with increasing areas of water bodies within the landscape plot. Richness of total species was also greater in the landscapes characterized by larger areas of water bodies, high proportion of noncrop vegetation, high compositional heterogeneity of noncrop habitat, or small field patches (high crop configurational heterogeneity). Crop compositional heterogeneity did not show significant effects neither on the whole community (all birds) nor on any of the three functional groups considered. These findings suggest that total bird diversity and some functional groups, especially woodland species, would benefit from increases in the proportion of noncrop features such as woody vegetation and water bodies as well as compositional heterogeneity of noncrop features within landscape.     

Regional distribution patterns predict bird occurrence in Mediterranean cropland afforestations

Part of the abandoned cropland in Mediterranean landscapes is being subjected to afforestation dominated by pines. Here we simultaneously evaluate the effect of three categories of factors as predictors of the interspecific variation in bird habitat occupancy of fragmented afforestations, namely regional distribution, habitat preferences, and life-history traits of species. We use the “natural experiment” that highly fragmented pine plantations of central Spain represent due to the prevailing pattern of land ownership of small properties. Many species with marked habitat preferences for woodland habitats were very scarce or were never recorded in this novel habitat within a matrix of deforested agricultural landscape. Interspecific variability in occurrence was mainly explained by regional distribution patterns: occurrence was significantly and positively associated with the proportion of occupied 10 × 10 UTM km squares around the study area, habitat breadth, and population trend of species in the period 1998–2011. It was also positively associated with regional occupancy of mature and large pine plantations. Other predictor variables related to habitat preferences (for woodland, agricultural and urban habitats) or life-history traits (migratory strategy, body mass, and clutch size) were unrelated to the occurrence of species. Thus, small man-made pinewood islands funded by the Common Agrarian Policy within a landscape dominated by Mediterranean agricultural habitats only capture widespread and habitat generalist avian species with increasing population trends, not contributing to enhance truly woodland species.     

Conserving natural enemies with flowering plants: Estimating floral attractiveness to parasitic Hymenoptera and attraction’s relationship to flower and plant morphology

Flowering plants in agricultural landscapes can provide ecological services, such as nectar-provision for adult parasitic Hymenoptera. Various flowering native, introduced/established and cultivated potted plants were used to bait interception traps along the wooded margins of fields planted seasonally with either feed-corn or rye. Depending on circumstances, controls consisted of traps baited with the same species of plant without flowers, a pot/area without plants, or both. In most cases pots were rotated among trap-sites. Of the 19 plant species tested, 10 captured significantly more summed ichneumonoids and chalcidoids, seven more Braconidae, two more Ichneumonidae and six more Chalcidoidea than controls. Among Braconidae, traps baited with certain plants captured significantly more individuals of specific subfamilies. “Attractive” and “unattractive” plant species tended to cluster in a principal components vector space constructed from plant morphological characteristics (flower width, flower depth, flower density and plant height). Flower width and plant floral-area (flower width² * flower density) were the variables that most often explained the variance in capture of the different parasitoid taxa. Our study identified particular plants that could be incorporated into regional conservation biological control programs to benefit parasitoid wasps In addition, the results indicate that morphological characteristics might help identify further suitable plant candidates for agricultural landscape modification.     

Cork-oak woodlands as key-habitats for biodiversity conservation in Mediterranean landscapes: a case study using rove and ground beetles (Coleoptera: Staphylinidae, Carabidae)

Land-use intensification in Mediterranean agro-forest systems became a pressure on biodiversity, concerning particularly the woodland sensitive species. In 2001, the effects of a land-use gradient from old-growth cork-oak forest to a homogeneous agricultural area were assessed using rove beetles as indicators in a Mediterranean landscape. The aim was to find which species were negatively affected by land-use intensification at the landscape level and whether they benefited from cork-oak patches occurring along the land-use gradient. A total of 3,196 rove beetles from 88 taxa were sampled from all landscape types. Agricultural area recorded significantly higher numbers of abundance and species richness in relation to the cork-oak mosaics, i.e. the old-growth forest and the managed agro-forest landscapes (montados). Moreover, 70% of rove beetle indicator species common enough to be tested by IndVal displayed their highest indicator value for agriculture, showing a lower number of woodland indicators in comparison to ground beetles. Nevertheless, one rove beetle taxon was considered a specialist of closed woodland mosaics while no specialist ground beetle was found for that landscape typology. Some rare rove beetle species were also important in typifying diversity patterns of old-growth cork-oak forests. Hence, future management in Mediterranean landscapes should take into account not only indicator species common enough to be tested by IndVal, but also rare and endemic species. Considering the added value of cork-oak woodland cover for sensitive rove and ground beetle diversity, the strengthening of cork-oak woodland connectivity seems to be a crucial management that is required in agricultural Mediterranean landscapes.     

The landscape context of cereal aphid-parasitoid interactions

Analyses at multiple spatial scales may show how important ecosystem services such as biological control are determined by processes acting on the landscape scale. We examined cereal aphid-parasitoid interactions in wheat fields in agricultural landscapes differing in structural complexity (32-100% arable land). Complex landscapes were associated with increased aphid mortality resulting from parasitism, but also with higher aphid colonization, thereby counterbalancing possible biological control by parasitoids and lastly resulting in similar aphid densities across landscapes. Thus, undisturbed perennial habitats appeared to enhance both pests and natural enemies. Analyses at multiple spatial scales (landscape sectors of 0.5-6 km diameter) showed that correlations between parasitism and percentage of arable land were significant at scales of 0.5-2 km, whereas aphid densities responded to percentage of arable land at scales of 1-6 km diameter. Hence, the higher trophic level populations appeared to be determined by smaller landscape sectors owing to dispersal limitation, showing the 'functional spatial scale' for species-specific landscape management.     

Species diversity and persistence in restored and remnant tallgrass prairies of North America: a function of species' life history, habitat type, or sampling bias?

1. The re-assembly of native animal communities in restored landscapes is a relatively unexplored phenomenon for many taxa. Specifically, ecologists lack the ability to generalize about how species traits, habitat size, habitat type (here, remnant prairie vs. restored grassland), and temporal variation interact to affect species diversity or species' persistence probabilities. 2. To investigate these relationships, moth communities from 10 prairie remnants and restorations were sampled over a 3-year interval and a combination of NMDS ordination, logistic regression, and repeated measures anova were used to test hypotheses regarding how life history variables and habitat characteristics determine the degree to which restored habitats develop a moth fauna similar to remnants. 3. Within sampling years, restored tallgrass prairies that were >or= 7 years old possessed lepidopteran species assemblages that were generally similar to those in prairie remnants. Community similarity, however, was driven by common moth species likely to also occur in the surrounding agricultural habitat. Species persistence was significantly influenced by a series of trait combinations identified using principal components analysis. Temporal variation independent of habitat type or patch size was the most significant determination of variation in species composition among sites. 4. These results suggest that lepidopteran persistence in restored landscapes is at least partially determined by species' life history attributes. The correlation between sampling year and species richness suggests that both weather effects on species voltinism and interannual differences in sampling bias may make it difficult for land managers to detect changes in species abundance following disturbance or habitat management. 5. Species may not necessarily possess specific life history traits that reduce extinction risk or enhance recolonization probabilities in the highly modified agricultural landscape of the Midwestern USA. Rather, voltinism, fecundity, body size, and host plant specialization may influence the ability of species to maintain populations in the greater agricultural landscape or to escape mass mortality following disturbances imposed by prairie management.